ownership structure, liquidity, and firm value...abstract a firm’s ownership structure influences...
TRANSCRIPT
Ownership structure, liquidity, and firm value:
Effects of the investment horizon
October 2010
Jun Uno* and Naoki Kamiyama†
* Waseda University, Graduate School of Finance, Accounting and Law.This research is supported by the Financial Service and Innovation Management Research Project, sponsored by the Ministry of Education, Culture, Sports, Science and Technology, FY2008. Correspondence to Jun Uno, 1-4-1 Nihombashi, Chuo-ku, Tokyo 103-0027, Japan; Tel. 81-3-3272-6798; fax 81-3-3272-6783; e-mail [email protected]. † Deutsche Securities Inc. We benefited from the comments and suggestions of Yakov Amihud, Marti G. Subrahmanyam, and Hung Wan Kat. We thank participants at Nippon Finance Association Annual Meeting 2009, The 22nd Australasian Finance and Banking Conference, Asian Finance Association Conference 2010, and 13th Conference of Swiss Society for Financial Market Research.
Abstract
A firm’s ownership structure influences both its liquidity and value. This paper investigate
the above relation by introducing a new measure of latent investment horizon, a weighted
average investment horizon computed from the firm’s ownership structure and the average
investment horizon of various investor categories. We find that the latent investment horizon
explains differences in liquidity and firm value among firms listed on the Tokyo Stock
Exchange. Empirical results indicate that the longer the investment horizon, the lower the
firm’s liquidity and value. In addition, concentrated ownerships by insider and cross-holding
shareholders can lead to inferior liquidity and firm value.
JEL Classification: G10, G32
Keywords: Ownership structure, liquidity, monitoring, corporate governance, market
microstructure
2
1. Introduction
This paper investigates how ownership structure is related to a firm’s market liquidity and
value. If the market liquidity of the firm’s shares declines due to concentrated ownership, the
value of the firm is expected to decrease. According to Bhide (1993) and Holmstrom and Tirole
(1993), illiquidity may result from increased asymmetric information. On the other hand,
large shareholder faces liquidity constraint to unwind her holdings so that she ought to
strengthen monitoring the firm’s management and contributes to increase a firm’s value.
There is a trade-off between illiquidity and level of corporate governance.
Maug (1998) and Kahn and Winton (1998) suggest that greater liquidity can be an
opportunity for large shareholders to increase their profit by monitoring the firm’s
management. They mention the case where a large shareholder chooses to buy more shares
when the firm’s performance is expected to improve as a result of monitoring activities. The
greater the liquidity, the more shares can be bought in the market due to lower transaction
costs. Thus, a higher concentration of ownership does not necessarily mean a trade-off
between corporate governance and illiquidity. This differs from past views such as in Bhide
(1993), where a stock’s high liquidity renders large shareholders less aggressive in their
monitoring and more likely to sell shares when they find poor performance of the firm’s
management.
Thus prior theoretical literature address the impact of concentrated ownership on a firm’s
market liquidity and value, but does not consider shareholders’ investment horizons which
differ substantially among shareholders. When a firm has many shareholders with a longer
(shorter) investment horizon, its market liquidity diminishes (increases). In this study we
3
investigate the relation between the average investment horizon of the entire ownership and
a firm’s liquidity as well as value that Amihud and Mendelson (1986) predicts.
Atkins and Dyl’s (1997) study on the relation between investor average holding period and
liquidity is similar to ours. They find a strong correlation between a stock’s share turnover
and bid-ask spread. Our paper differs from theirs in the following regards.
First, we use the latent investment horizon instead of turnover. Turnover, the authors’ proxy
for the shareholders’ investment horizon, is observed ex post and therefore deviates from the
ex ante average holding period of the firm’s ownership structure, because turnover is largely
affected by short-term trading activity and informational events such as quarterly reports
and takeover bids. Our paper focuses on how ownership structure affects stock liquidity. We
need a proxy for the ex ante average holding period of shareholders that is not computed from
a realization of mixed trading interests. Second, we use a liquidity measure such as Amihud’s
(2002) ILLIQ in addition to the bid-ask spread. ILLIQ is reflected by not only the bid-ask
spread but also market impact and thus represents a wider scope of liquidity, which is
relevant for both small individuals and large institutional investors as an indicator of
transaction cost.
For the ex ante investment horizon, this study calculates a weighted average investment
horizon of a firm’s shareholders for Japanese companies listed in the First and Second
Sections of the Tokyo Stock Exchange (TSE). We follow Mahanti et al. (2008), who were the
first to use latent liquidity to estimate transaction cost of corporate bonds. In their study, the
authors estimate the investment horizon of corporate bondholders using data from custodian
banks. In our case, however, there are no corresponding data available from custodian banks.
Therefore, we estimate the latent investment horizon of each stock from the ownership ratio
4
and the market wide investment horizon of investor category, such as foreigners, banks, and
individuals. The average investment horizon of each investor category is computed from the
aggregate amount of the holding and the total trading volume of TSE-listed companies.
By considering the average investment horizon of the ownership structure, we can address
cases in the real economy where, under the same concentrated ownership, a firm’s market
liquidity and value are higher because the weighted average investment horizon of its
shareholders is shorter than others. We show that a firm’s weighted average investment
horizon is highly correlated with market liquidity and firm value. We expect that the shorter
the investment horizon, the higher the liquidity and value of the firm. Additionally, there are
two distinct categories of large shareholders in Japan: foreigners and cross-holders (mochiai).
These investors are opposites in terms of investment horizon and monitoring management.
We shed some light on specific investor categories and whether the size of their presence
affects firm liquidity and value.
Our results are summarized as follows: (1) The longer the latent investment horizon, the
lower the liquidity. (2) Longer investment horizon leads to lower firm value. (3) Investor
category has a distinct effect on liquidity, with foreigners impacting positively and
cross-holders negatively. The results imply that ownership structure relates to a firm’s
liquidity and value. Our results indicate that the higher the proportion of short horizon
shareholders, the higher the firm’s liquidity and value. On the other hand, a firm with a high
proportion of infrequent investors who do not monitor management and facilitate the
entrenchment of current management results in lower firm liquidity and value. Based upon
these results, strengthening cross-holding is an inferior corporate policy, since it ultimately
impairs the liquidity of the entire market.
5
The remainder of this paper is organized as follows. Chapter 2 proposes a new approach to the
relation between ownership and liquidity, with a brief overview of previous work. The data
and sample stocks are described in Chapter 3. Chapter 4 presents our empirical results and
Chapter 5 gives our conclusions.
2. A New Approach and Hypotheses
2.1 Prior research
A firm’s ownership concentration influences its liquidity and value. We argue that the manner
in which ownership structure affects liquidity depends upon a weighted average investment
horizon of the firm’s shareholders. If the average investment horizon of the firm is longer,
then the illiquidity of its shares is more severe.
Amihud and Mendelson (1986) model predicts that illiquid stocks are owned by investors with
longer investment horizon. If these investors do not actively monitor a firm’s management
and they do not trade, the firm’s share liquidity remains low. As suggested by Maug (1998),
and Kahn and Winton (1998), if a firm’s shareholders commit to monitoring management,
they trade frequently to maximize their profits from their private information and thus
contribute to improve the stock’s liquidity.
Bhide (1993) and Holmstrom and Tirole (1993) demonstrate a negative correlation between
ownership concentration and firm liquidity. When the founding shareholder owns all of a
firm’s shares, there is no liquidity. If the founding shareholder sells off a small part of the
shares, liquidity improves but monitoring incentives are decreased gradually. This is due to a
free rider problem in which minority shareholders enjoy the monitoring efforts of a large
shareholder. On the other hand, the informational advantage of large shareholders who
6
commit to monitoring increases information asymmetry among investors, causing lower
liquidity. Thus, a trade-off exists between liquidity and monitoring. Neither of the authors
considers the investment horizon of the firm’s shareholders.
Kahn and Winton (1998) and Maug (1998) argue that large shareholders might not release
their ownership when market liquidity is high. While a large shareholder continues
monitoring management to improve the firm’s value, more shares can be bought to maximize
profit. The decision on how many shares to add depends upon market liquidity or transaction
costs. The authors also point out that for monitoring to remain profitable, it is crucial that the
monitoring information be accurate. They do not, however, examine the investment horizon of
large shareholders.
A small number of empirical studies have been carried out on the trade-off between liquidity
and corporate governance. Gaspar and Massa (2007) empirically examine the trade-off
between monitoring and liquidity. They show that informed ownership improves governance
and induces value-enhanced decisions, but reduces liquidity due to increased adverse
selection cost. Rubin (2007) finds that liquidity is positively correlated to total institutional
holdings but negatively correlated to institutional block holdings. The level of institutional
holdings proxies for trading activity, and the concentration of ownership, such as block
holdings, proxies for adverse selection costs. Sarin and et al. (2000) analyze ownership
concentration by insiders and by institutional investors. They report decreased liquidity in
both cases: by insiders from increased asymmetric information and by institutional investors
from inventory costs. Garvey and Swan (2002) empirically verify Holmstrom and Tirole’s
(1993) hypothesis with a sample of 1,500 U.S. companies and report that high liquidity has a
positive impact on shareholder value.
7
Atkins and Dyl’s (1997) study, which is similar to ours, on the relation between investor
average holding period and the bid-ask spread for NASDAQ and New York Stock Exchange
stocks finds strong evidence that the turnover ratio, computed by dividing a firm’s number of
outstanding shares by its annual trading volume, is related to the bid-ask spread. The
authors do not consider ownership structure, so they do not examine whether higher turnover
is due to a firm’s ownership structure or not. It is important to distinguish latent liquidity
from trading volume. Therefore, we introduce a new measure of the weighted average
investment horizon to determine the level of liquidity for individual stocks.
Effects from investment horizon of institutional investors have mixed results. Yan and Zhang
(2007) conclude that short-term institutions are better informed based upon the facts that
short-term institution’s trading is positively related to future stock return and earnings
surprises. On the other hand, Gaspar et al. (2005) and Chen et al. (2007) report that higher
holdings by long-term investors are associated with improvement of post-event performance
in case of takeover and merger & acquisitions.
With regards to foreign investors, Tesar and Werner (1995) find that the turnover rate on
equity held by non-residents is higher than the overall turnover rate on the domestic market.
Foreigners respond to changes in economic conditions by making frequent and sizable shifts
in their holdings of foreign securities, even though much of this activity has little impact on
net investment position. According to Nitta’s (2000) analysis of data from 1988 to 1997, a
positive correlation exists between the ratio of foreign ownership and management
performance measures such as return on equity (ROE), but a negative correlation is observed
in the case of the cross-holding ownership ratio and management performance.
8
In Japan, foreigns and cross-holding shareholders are important constituents in ownership
structure. Foreigners typically have a shorter investment horizon and are apt to monitor
management, they expect to contribute improvement of liquidity and value, whereas
cross-holding shareholders typically have a longer investment horizon and are less active in
monitoring management. We expect that the larger the percentage of foreign ownership, the
higher the firm value; on the other hand, the larger the percentage of cross-holding owners,
the lower the firm value. The Japanese market provides an ideal data set to test the
previously proposed hypothesis.
2.2 Hypotheses
We test the following hypotheses with respect to the relation of ownership structure, liquidity
and firm value.
H1: The longer the investment horizon, the lower the liquidity. If a firm has many long-term
investors, they will trade infrequently such that the liquidity of the stock will remain low.
H2: The higher the ownership concentration, the lower the liquidity. Ownership concentration
is to strengthen information asymmetry, which increases transaction costs, reduce liquidity1.
The proxy of concentration is top30 ownership. We separate the top30 ownership into insider
and non-insider portion to investigate the effect of which is greater.
H3: We test whether investor category who owns large amount of shares affects size and
direction of impact on liquidity. When high cross-holding ratio is associated with high top30
ratio, we expect an increase in illiquidity, whereas when high foreign ownership ratio is
associated with high top30 ratio, we expect decrease in illiquidity. Because foreigners are 1 A definition of Insider is shares held by insiders within top thirty shareholders. That of Top30 extracts insiders’ share ownership from that of top thirty shareholders.
9
short horizon investors and sensitive to management entrenchment, their large presence
reduces negative impact on liquidity caused by concentration itself. Cross-holders are
opposite to foreigners with respect to investment horizon and monitoring management.
H4: For firm value, the shorter the horizon, the higher the value due to positive liquidity
effect.
H5: Presence of large shareholders creates tradeoff between monitoring and liquidity because
there are two countervailing effects, negative effect from asymmetric information and positive
effect from monitoring. Which factor has larger effect on corporate value is an empirical
question. Considering investor categories who own large proportion of a firm and whose
investment horizon is significantly different, we may point out factors strongly related to firm
value.
3. The Data
We use four years of data on First and Second Section companies on the TSE, from 2004 to
2007. The sample includes 1,657 to 1,686 stocks for which liquidity and cross-holding data are
fully available. As shown in FigureⅠ, the cross-holding structure of Japanese companies
changed rapidly during this period. The Program for Financial Revival implemented in 2002
accelerated the unwinding of cross-holding by banks, while the presence of foreign investors
rose.
10
3.1 Investment horizon
A firm’s investment horizon is estimated from two sources of the data. First, the investment
horizons of an investor group are computed from the aggregated market data provided by the
TSE. Each investor group’s investment horizon is the total trading volume during the year
divided by the average portfolio market value at the start and end of each year. The data
source for stock ownership by investor group is the TSE’s Share Ownership Survey. The
trading amount of each investor group is compiled by the TSE and published annually in its
Investment Trends by Investor Group.2 There are four investor categories used at this stage
of the estimation; foreigners, individuals, non-financial corporations, and a group of financial
institutions (trust banks, insurance companies, and banks).3 The following is the equation for
an investor group j’s investment horizon in year t:
(1) jt
jt
jt
jt
volumetradingtotal
valuemarketportfoliovaluemarketportfolioHorizonInvestment
yen 2
1) ( 1
(Table 1 around here)
Table 1 shows the average investment horizon for each investor group at the aggregate level.
A non-financial corporation’s investment horizon (Horizon) is 5 to 7 years and an insurance
company’s Horizon is 15 to 21 years, which is the most inactive among six investor categories
in Table 1. The Horizon of foreigners is 0.2 to 0.4 year, which means they trade two to five
2 The original monthly data frequency was converted into fiscal years. The ownership survey covers seven local markets in Japan, whereas trading volume covers three major markets. The volume on four local exchanges account for about 0.01% of total trading volume in the entire market. 3 In the classification of financial institutions, the investor categories in annual reports do not match the three subgroups available in the TSE statistics, such as commercial banks, insurance companies, and trust banks. Therefore, we take an average of the investment horizons for commercial banks, insurance companies, and trust banks as the investment horizon for financial institutions when computing the investment horizons of the TSE-listed companies.
11
times a year. The Horizon of individuals is 0.3 to 0.5 year. The finding that foreigners’ Horizon
is shorter than that of domestic investors is consistent with Tesar and Werner (1995). The
authors report turnover rates in domestic equity held by foreign residents for five major
countries—including the United States, the United Kingdom, and Japan—and find that
foreign investors transact at a significantly higher rate than domestic investors.
Next, we compute Horizon for firm k in year t as follow:
(2)
4
1,
j
jt
jtk
kt HorizonwHorizon
where j represents one of four investor categories, foreigners, individuals, non-financial
corporations, and financial institutions; is a firm j’s ownership ratio obtained from the
company’s annual report; and is the market-wide investment horizon by investor
group. The data source for ownership structure was QUICK’s AMSUS. Thus, the investment
horizons for all listed firms are estimated based upon companies’ ownership structure
jtkw ,
ktHorizon
4.
4 The following example illustrates the computation of equation (2). In 2007 Sony Corporation’s
ownership ratios of foreigners, individuals, non-financial corporations, and financial
institutions were 50.73, 23.11, 3.10, and 21.57%, respectively. Using the and , we
compute the weighted average of Sony’s investment horizon, which is about 2.8 years:
jtkw ,
jtHorizon
jtkw ,
year2.82.391970.209870.086200.11008
21.57%15.650)16.659(0.9593
1 3.1%6.7723.11%0.37350.73%0.217
HorizonofaverageweightedA
12
We use the NEEDS database as an additional source of data on large individual shareholders
such as the founder (owner) of a firm and, national and local governments on the top 30
shareholders list. We use these data to adjust the above computation. Since neither a founder
(owner) nor a government trade like ordinary individual investors, we assign the longest
investment horizon among six sub-investor categories in the same year to their ownership
category.
3.2 Ownership structure
3.2.1 Ownership concentration by insiders and non-insiders
Our measure of ownership concentration is the sum of the top 30 shareholders’ holdings
divided by the total number of shares outstanding. In addition, we partition block-holdings
into its two major components, insiders’ and non-insiders’ equity holdings 5 . Higher
concentration by insiders may be a signal of weak governance, whereas concentration by
non-insider serves as a proxy for the probability that informed investors participate in the
market. The concentration by non-insiders strengthens the severity of adverse selection costs.
3.2.2 Investor category ownership ratio
Equity holding by an investor category is computed as the number of shares owned by each
investor category divided by the total number of shares outstanding. Foreign and individuals
investor are used in a regression analysis.
5 Rubin (2007) uses insider and institutional block holdings to measure concentration. Hartzell and Starks (2003) use a measure that is the top five non-insider institutional investors’ holdings divided by the total number of institutional holdings.
13
3.2.3 Cross-holding
Our measure of cross-holding shareholders is the Mochiai holding ratio of each stock,
estimated by the NLI Research Institute. The Mochiai holding occurs when two listed
companies mutually hold shares, confirmed through disclosed materials.6 An average Mochiai
ratio is around 9% and the maximum is above 50%.7
3.3 Liquidity
Total trading cost consists of bid-ask spread and market impact cost. Goyenko et.al.(2009)
suggests that reliable proxy for spread and market impact are different. We select Amihud
(2002)’s illiquidity measure8 for the proxy of market impact and quoted spread9 for the proxy
of adverse selection cost.
3.3.1 Market impact measure
Amihud measure is the monthly average of the absolute daily return divided by the daily yen
volume. We eliminate stocks traded less than 10 days per month. Here, ILLIQ shows the
relation between price change and volume; it is a rough estimate of spread plus market
impact cost. We use the average relative illiquidity (RILLIQ) for each fiscal year:
(3)
iyD
t ivyd
iyd
iyiy VOL
R
DILLIQ
1
1
6 Alternative measure is the Antei (=stable) holding, where the ownership of a firm’s shares by banks and life insurance companies is disclosed but the firm’s holdings of counterparty shares cannot be confirmed by disclosed materials. Our choice of cross-holding measure does not affect the empirical results. Refer to Nitta (2002) for details. 7 Nitta (2002) shows that the higher the cross-holding ratio, the worse the various managerial indices, such as the ROE. 8 This measure is one of best measure as a proxy for market impact, according to Goyenko et.al.(2009) 9 Quoted spread is widely used as a proxy for adverse selection costs such as Atkins and Dyl (1997) and Rubin (2007).
14
(4)
yN
tiy
iyiy
ILLIQN
ILLIQRILLIQ
1
1
Here, RILLIQ is adjusted for market-wide liquidity changes and converted into a natural
logarithm10. According to the literature, liquidity has a commonality (Chordia et al. 2000),
and its market-wide fluctuation has differential impacts on the price of individual stocks
(Amihud 2002). Thus, RILLIQ is adjusted for a time series variation of market-wide liquidity
changes.
3.3.2 Bid-ask spread
Quoted spread (SPRD) is defined as the difference between lowest ask price and highest bid
price divided by the mid-price of the quotes. SPRDs are calculated every time when best ask
and/or bid changes, we compute a time-weighted average spreads for stock j on day t, and
then average them over a year. We exclude any quotes before the opening price.
(5) 2)(
)(
,,
,,,
tjtj
tjtjtj
BestBidBestAsk
BestBidBestAskSPRD .
3.4 Firm value
Our proxy for firm value is so-called Tobin’s Q defined as follow:
(6) 2)debt earinginterest_b + capital total(
)tearing_debinterest_b + uemarket_valaggregate_(
tj,tj,
tj,tj,, tjQRATIO .
10 The impact of a change in market-wide liquidity on ILLIQ for individual stocks is not uniform: It
disproportionately affects low-liquidity stocks (see Amihud 2002 and Chordia et al. 2000).
15
The financial data for calculating this are from World Scope. When we compute Tobin’s q, the
cross-holding-adjusted aggregate market value is used. 11 Table 2 shows the summary
statistics of these variables.
3.5 Cross-sectional correlations
Table 3(A) shows the cross-sectional correlations among the variables. Horizon is positively
correlated with the logarithm of market capitalization (0.3) and negatively correlated with the
share turnover ratio (0.11), as shown in Table 3(A). The correlation with the ownership ratios
of Mochiai are positive (0.3), meaning that the longer Horizon, the higher the Mochiai ratio.
On the other hand, there is almost no correlation between Horizon and Foreigner (see Table
3(B)).
(Table 3 around here)
4. Empirical Analyses
4.1 Ownership structure and illiquidity
First, we test the relation between investment horizon and illiquidity. Illiquidity and
ownership structure may be simultaneously determined. All variables are most likely
endogenous and the estimates based on panel least squares are biased and inconsistent.
Considering these problems, we use a two-stage estimation method with instrumental
variables to obtain a consistent estimate.12 Our basic regression equation is
(7) tjtjtjtjtj sriabledControlValestionVariabcConcentrabHorizonayIlliquidit ,,,,, )(
11 Kobayashi (1990) points out that the Mochiai portion should be subtracted from the aggregate market value to calculate Tobin's q. 12 Woodridge (2002) Chapter 10 -11.
16
A proxy for market illiquidity is RILLIQ and bid ask spread (SPRD). For concentration
measures, TOP30, and a set of INSIDER and NON-INSIDER. are used interchangeably. We
also test the interaction between NON-INSIDER and investor category holdings). The natural
logarithm of a firm’s market value (Log_Size) is added as control variable.13 Our instrumental
variables are the lagged explanatory variables. Heteroskedasticity is corrected by White
diagonal standard errors and covariance corrections.
In model1 of Table4, the coefficient of Horizon and Top30 are positive and significant at the
1% level. It means that the longer the investment horizon, the lower the liquidity, and the
higher the concentration, the lower the liquidity. These findings are consistent with H1 and
H2.
In model2, TOP30 variable is separated into insider and non-insider portion of concentration,
INSIDER and NON-INSIDER respectively. The result shows that both variables have
positive correlation with illiquidity, the coefficient of insider concentration is much larger
than that of non-insider concentration (2.248 vs. 0.379). Insider’s ownership concentration
have a bigger negative impact on liquidity than non-insider’s. It indicates when insiders own
large portion of the company’s shares outstanding, liquidity provided by existing shareholders
are limited to cause large market impact.
Model3 examines whether large presence of specific investor category has relation with
liquidity under the concentrated ownership. We are interested here in the magnitude and
direction of the investor category’s influence. We insert cross-term variables such that
(NON-INSIDER x holding ratio of investor category such as Foreign, Indiv, and Crosshld ) as
explanatory variables. The result shows that Foreign and Indiv are insignificant, but 13 Amihud (2002) shows a high negative correlation (-0.614) between firm size and RILLIQ. In our case the correlation is -0.57.
17
Crosshld has a significant positive coefficient with ILLIQ. It means that only case where the
high non-insider concentration is associated with high cross-holding ratio, illiquidity
increases. The result indicates not only how large the concentration but also what types of
investors own such large proportion. Cross-holders own shares not for pure investment
purpose, does not trade, and provide less commitment to monitoring, so that other market
participants see these factor negatively for liquidity. In case of foreigner and individuals, their
short horizon investment style mitigate negative effect from asymmetric information on
liquidity.
Next we use SPRD% to run regression equation (8), where SPRD% is the percentage of
bid-ask spread in year t for firm j. Since the severity of asymmetric information affects the
size of the spread, we expect that the higher the TOP30, the wider the spread. We include
number of trades per day (Trade), relative tick size (Tic/Price)14 as control variables.
In table 5, the coefficient of TOP30 is positive and significant, however that of HORIZON is
insignificant. As Goyenko et.al. (2009) suggest that the results support notion that bid-ask
spread is directly related to adverse selection cost and ownership concentration is more
important than investment horizon. In model2 both of INSIDER and NON-INSIDER show
positive correlation with SPRD%, the coefficient of INSIDER is larger than that of
NON-INSIDER (0.4365 vs. 0.2079). It means that higher concentrations by insiders as well as
non-insiders increase the bid-ask spread which is contrary to Rubin (2007) which reports
negative correlation between insider holdings and bid-ask spread. Our result indicates that
insider holdings are related to asymmetric information.
14 Unlike the New York Stock Exchange, the TSE uses a tick size that is a step function of share price. In the sample period of 2004 and 2007, the tick size is ¥1 for stocks priced below ¥2000, ¥5 for stocks priced between ¥2,001 and ¥3,000, ¥10 for stocks priced between 3,001 and ¥30,000. For further detail, see TSE(2010)
18
Model3 examines investor categories’ effect on liquidity. The coefficients of Foreign x
NON-INSIDER and Crosshld x NON-INSIDER are positive, but that of Indiv x
NON-INSIDER is negative. Large presence of cross-holders and foreigners increase adverse
selection cost, while individuals mitigate it. It shows that foreigners have superior
information which increases information asymmetry among investors.
4.2. Ownership, liquidity, and their effects on firm value
We extend an analysis to the relation among investment horizon, ownership concentration
and firm value. Tobin’s q (QRATIO) is used as a proxy for the firm’s market valuation. The
basic regression equation is
(8) ariablesdControllVesionVariablcConcetratriablebHorizonVaaQRATIO
where horizon variable and concentration variables are same as equation (8) and the debt
asset ratio (Debtasset) and the profit growth rate (Growth) are included as control variables.
Debtasset is calculated as total debt divided by total assets, and Growth is calculated as the
one-year growth rate of the ordinary profit which is equivalent to income before extraordinary
items and taxes. In order to avoid endogeneity problem, we use the two-stage estimation with
instrumental variables to obtain a consistent estimate. Our instrumental variables are the
lagged explanatory variables. A panel regression analysis is carried out according to the result
of the Hausman test which rejects the random effect model for both time and cross section,
and then the fixed time-effect model is selected.15
15 The Hausman test rejects the random effect model at p = 0.0099 with equation (6) and at p =
0.0012 with equation (7).
19
The estimated results are shown in Table 6. HORIZON is negatively correlated with the firm
value. It means that longer investment horizon of shareholders has negative impact on firm
value. This is interpreted as illiquidity effect. TOP30 is positively correlated with the firm
value. The result supports the notion that block-holders influence corporate governance and
improve performance. On the issue of trade-off between monitoring and liquidity (H5), this
result indicates that positive impact of monitoring activity is larger than negative impact of
illiquidity caused by concentration16. This is a unique and important finding of this study in
the relation of corporate governance and market microstructure.
In model2 INSIDER and NON-INSIDER are significant. The coefficient of NON-INSIDER is
slightly larger than that of INSIDER (0.0122 vs. 0.0100, respectively). We expect pressure
from non-insider block holder is larger than that from insiders, but the difference is not as
large as expected.
Model3 examine whether specific investor category influences to firm value. The variable of
(NON-INSIDER x investor category such as Foreign, Indiv, and Crosshld ) shows differential
effects on firm value. As we expect, Foreign has a significantly positive relation with Q-ratio,
but Crosshld has a significantly negative relation. The results support Leleux, Vermaelen,
and Banerjee (1995) that analyze the impact of large block-holders on firm performance in
France and find that the identity of the block-holders is crucial.
These findings indicate that ownership structure and its composition affect firm’s valuation.
If there is high crossholding relationship, market participants judge the firm to be lax in
uidity
t 16 As a related study, Gaspar and Massa (2007) examine the trade-off between ownership, liqand firm value. They find that the effect of ownership on Tobin’s Q is not statistically significanafter controlling the endogeneity of ownership by IV estimate.
20
corporate governance and discount the firm’s value. Ignorance of corporate governance
severely deteriorates firm’s valuation as well as liquidity.
(Table 6 around here)
4.3 Robustness check,
As a robustness check, we investigate how changes in HORIZON and TOP30 affect liquidity
measures. We will see the lagged relationship between them. When a company’s weighted
average of investment horizon or concentration of ownership changes, existing and new
shareholders must buy or sell shares in the market. It affects our measures of liquidity and
firm value on the same year due to large movement of position made by institutions and
foreign investors. For our purpose to confirm robustness of the relation among ownership,
liquidity and firm value, we should ask the following question. When a firm’s ownership
concentration declines, what happens to liquidity and firm value on the following year? Thus
the equation (9) is estimated with three different dependent variables.
(9)
tjtjtjtjtj riablesControllVabTOPbHORIZONbaValueyIlliquiditVariables ,,31,21,1, 30),(
Dependent variables ⊿Variables a change of ILLIQ, SPRD, and QRATIO. In equation (9)
where ⊿ILLIQ is the change in ILLIQ for stock j from year t-1 to year t. ⊿SPRD% is the
change in percent bid-ask spread of stock j from year t-1 to year t. ⊿QRATIO is the change
in QRATIO of stock j from year t-1 to year t. In addition to ⊿HORIZON and ⊿TOP30, we
21
include ⊿Market_ILLIQ for ⊿ILLIQ, ⊿Tic÷⊿Price
17 for⊿SPRD%, and ⊿Debtasset and
⊿Growth to as control variables,
The results are presented in Table 7. With respect to ⊿ILLIQ , larger ⊿HORIZON causes
greater deterioration on liquidity. Increasing insider ownership is positively correlated with
illiquidity on the next year, increasing non-insider ownership has the opposite effect on
illiquidity.
With respect to ⊿SPRD%, ⊿HORIZON has negative coefficient. It means that increased
short-term shareholders increases adverse selection risk for other market participants. When
insider and non-insider concentration increases, the bid-ask spread widens as well.
With respect to ⊿QRATIO, ⊿Insider and ⊿Non-Insider at t-1 have positive impact on
firm value. This supports the notion that higher concentration improves governance and firm
value. Maug (1998), for instance, suggests that the characteristics of the large shareholders,
such as institutions, and their organizational structure are potentially important aspects for
the success of monitoring activities.⊿HORIZON have negative but insignificant relation with
⊿QRATIO .
In summary, investment horizon and ownership concentration are important factors that
influence both liquidity and value of firm’s share. When they change, there are associated
changes in liquidity and firm value.
(Table 7 around here)
17 This is an adjustment of the effects of the TSE’s step-wise tick table.
22
5. Conclusions
We empirically show that the weighted average investment horizon of a firm’s ownership
structure affects its liquidity. In addition, investment horizon and monitoring management
influence a firm’s value.
Our empirical results are summarized in the following four points.
(i) The latent investment horizon relates to a firm’s market liquidity. The latent investment
horizon is computed from the ownership structure and the average investment horizon of the
investor categories. The longer the latent investment horizon, the lower the liquidity.
(ii) Concentrated ownership has negative impact on liquidity. The investor categories
considered in this study are insider, non-insiders, foreign, cross-holding and individuals
When insiders own large proportion of the shares, negative effect is larger compared to the
case where non-insiders own large proportion of the shares. Effects from concentrated
ownership depend upon who owns them. Under the concentrated ownership, the higher the
cross-holding, the lower liquidity. The greater foreign ownership widens SPRD% but not
ILLIQ. This indicates asymmetric information effects is more important for bid-ask spread
measures.
(iii) The investment horizon influence firm value. A shorter investment horizon has a positive
impact on firm value.
(iv) Ownership structure affects firm value by signaling a firm’s corporate governance status.
The higher the cross-holding, the lower the firm value. Monitoring by foreign investors,
however, contribute improvement of market liquidity as well as the firm’s valuation directly.
23
The results support the notion that the composition and investment horizon of ownership
structure influence market liquidity and firm value. It indicates that a weighted average
investment horizon of the firm’s shareholders is an important characteristics of firm’s
ownership structure with respect to liquidity and value.
24
References
Amihud, Y. 2002. Illiquidity and stock returns: Cross-section and time-series effects. Journal
of Financial Markets 5, 31-56.
Amihud, Y., and Mendelson, H. 1986. Asset pricing and the bid–ask spread. Journal of
Financial Economics 17, 223-249
Amihud, Y., and Mendelson, H. 2008. Liquidity, the value of the firm, and corporate finance.
Journal of Applied Corporate Finance 20-2 32-45.
Amihud, Y., Mendelson, H., and Uno, J. 1999. Number of shareholders and stock price:
Evidence from Japan. Journal of Finance 54-3, 1169-1184.
Atkins, A., and Dyl, E. 1997. Transactions costs and holding periods for common stocks.
Journal of Finance 52-1, 309-325.
Bhide, A. 1993. The hidden costs of stock market liquidity. Journal of Financial Economics 34,
31-51.
Chen, X., Harford J., and Li, K. 2007. Monitoring: Which institutions matter?, Journal of
Financial Economics 86, 279-305.
Chordia, T., Roll, R. and Subrahmanyam, A. 2000. Commonality in liquidity, Journal of
Financial Economics 56-1, 3-28.
Davies, J. R., Hillier, D., and McColgan, P. 2005. Ownership structure, managerial behavior
and corporate value. Journal of Corporate Finance 11, 645-660.
Dahlquist, M., and Robertsson, G. 2001. Direct foreign ownership, institutional investors, and
firm characteristics. Journal of Financial Economics 59 413-440.
Gaspar, J., Massa M., and Matos, P., 2005. Shareholder investment horizons and the market
for corporate contorol, Journal of Financial Economics 76, 135-165.
Gaspar, J., and Massa M, 2007. Local ownership as private information: Evidence on the
monitoring-liquidity trade-off, Journal of Financial Economics 83, 751-792.
Glosten, L., and Harris, L., 1988. Estimating the components of the bid/ask spread, Journal of
Financial Economics 21-1, 123-142.
Glosten, L., and Milgrom, P., 1985. Bid, ask and transaction prices in a specialist market with
heterogeneously informed traders, Journal of Financial Economics 14-1, 71-100.
Goyenko, R.Y., Holden C.W., and Trzcinka C.A., 2009. Do liquidity measures measure liquidity,
Journal of Financial Economics 92, 153-181.
Hartzell J., and Starks L., 2003. Institutional Investors and Executive Compensation,
Journal of Finance 58-6, 2351-2374.
25
Hasbrouck J. 2009. Trading Costs and Return for U.S. Equities: Estimating Effective Costs
from Daily Data. Journal of Finance 64-3, 1445-1477.
Hausman, L. 1982. Specification tests in econometrics. Econometrica 46-6, 909-938.
Holmstrom, B., and Tirole, J. 1993. Market liquidity and performance monitoring. Journal of
Political Economy 101-4, 678-709.
Hotchkiss, E., and Strickland, D. 2003. Does shareholder composition matter? Evidence from
the market reaction to corporate earnings announcements. Journal of Finance 58-4,
1469-1498.
Huang, R., Stoll, H., 1997. The components of the bid-ask spread: A general approach, The
Review of Financial Studies 10-4, 995-1034, 1997
Kahn, C., and Winton, A. 1998. Ownership structure, speculation, and shareholder
intervention. Journal of Finance 53-1, 99-129.
Kang, J., and Stulz, R. M. 1997. Why is there a home bias? An analysis of foreign portfolio
equity ownership in Japan. Journal of Financial Economics 46, 3-28
King, M., and Segal, D. 2008. The long-term effects of cross-listing, investor recognition, and
ownership structure on valuation. Review of Financial Studies 22-6, 2393-2404
Leleux, B., Vermaelen, T., and Banerjee, S., 1995, Large shareholdings and corporate control:
An analysis of stake purchases by French holding companies, Working paper, INSEAD
Mahanti, S., Nashikkar, A., Subrahmanyam, M., Chacko, G., and Mallik, G. 2008. Latent
liquidity: A new measure of liquidity, with an application to corporate bonds. Journal of
Financial Economics 88-2, 272-298.
Maug, E. 1998. Large shareholders as monitors: Is there a trade-off between liquidity and
control? Journal of Finance 53-1, 65-98.
Morck, R., Nakamura, M., and Shivdasani, A. 2000. Banks, ownership structure, and firm
value in Japan. Journal of Business, 73-4, 539-567
Nitta, K. 2000. Cross-sharing and enterprise management, positive analyses concerning
influence of ownership structure (in Japanese). Security Analysts Journal 38-2, 72-93
Rubin, A. 2007. Ownership level, ownership concentration and liquidity. Journal of Financial
Markets 10, 219-248.
Sarin, A., Shastri, K., and Shastri, K., 2000. Ownership structure and stock market liquidity.
Working paper, University of Pittsburgh, Pittsburgh, Pennsylvania.
26
Tesar, L., and Werner, I. M., 1995. Home bias and high turnover. Journal of International
Money and Finance, 14-4 467-492.
Tokyo Stock Exchange, 2010. Guide to TSE Trading Methodology
Woodridge, J., 2002. Econometric Analysis of Cross Section and Panel Data. MIT Press,
Cambridge, Massachusetts.
Yan, X., and Zhang Z. 2009. Institutional Investors and Equity Returns: Are Short-term
Institutions Better Informed? Review of Financial Studies 22-2, 893-924
27
Figure 1. Ownership ratio trends by investor category. This figure shows the trend in the ownership structures of Japanese companies for the period
2002 to 2007. The foreigner ownership ratio is that at the end of the fiscal year as reported by
QUICK-AMSUS. The antei (stabilizing holding) and mochiai ratios of each stock are
estimated by the NLI Research Institute. A mochiai holding occurs when two listed companies
mutually hold shares, confirmed through disclosed materials. An antei holding refers to cases
where the ownership of a firm’s shares by banks and life insurance companies is disclosed but
the firm’s holdings of the counterparty shares cannot be confirmed by disclosed materials. The
mochiai figures are included in the antei holding ratio. The sample includes stocks in the
First and Second Sections of the TSE.
0%
5%
10%
15%
20%
25%
30%
F2002 F2003 F2004 F2005 F2006 F2007
Foreigner
AnteiMochiai
28
Table 1. Average investment horizon by investor category.
The investment horizon by investor category is computed from the annual turnover ratio, that
is, the average of the aggregated market value of an investor category’s ownership at the start
and end of each year divided by the total trading amount of the investor category during the
year.
FiscalYear
Foreigner IndividualsNon-financial
CorporatesTrust Banks Insurance Banks
2004 0.414 0.531 7.488 1.238 15.451 11.5152005 0.331 0.304 5.711 1.065 18.845 13.1432006 0.303 0.393 6.615 1.206 21.274 17.5682007 0.219 0.373 6.773 0.959 16.659 15.650
Unit: year.
Source: TSE’s Share Ownership Survey,. and the Investment Trends by Investor Category.
29
Table 2. Summary statistics.
This table shows the summary statistics of the explanatory variables. Here, HORIZON is a
weighted average of the investment horizons of sample firms, TOP30 is the sum of the top 30
shareholders’ holdings divided by the total number of shares outstanding, QRATIO is Tobin’s q
= (aggregate market value + interest-bearing debt)/(total capital + interest-bearing debt), where
the aggregate market value is adjusted for the cross-holding proportion of ownership. Crosshld,
Foreign, Indiv are the percentages of ownership of firm j’s investor categories (cross-holding,
foreigner and individuals) at the end of fiscal year t
FY Mean Median Maximum Minimum Std Dev. #Obs.HORIZON ALL 4.81 4.89 10.12 0.28 1.34 6,695
(Year) 2004 4.72 4.82 8.00 0.28 1.24 1,6722005 5.29 5.40 10.12 0.43 1.45 1,6802006 5.29 5.40 10.10 0.43 1.45 1,6802007 4.66 4.83 8.63 0.31 1.30 1,686
TOP30 ALL 50.84 50.60 91.38 0.73 15.74 6,604(%) 2004 51.16 51.11 91.20 4.96 15.64 1,633
2005 51.40 51.08 91.38 3.89 15.71 1,6492006 50.46 50.10 90.19 2.59 15.66 1,6582007 50.36 50.05 88.51 0.73 15.92 1,664
QRATIO ALL 1.25 1.05 18.91 -9.33 0.88 6,6852004 1.22 1.05 12.89 0.22 0.77 1,6572005 1.48 1.22 18.91 -9.33 1.14 1,6672006 1.29 1.10 11.90 -7.50 0.87 1,6762007 1.01 0.89 9.26 0.24 0.59 1,685
Crosshld ALL 9.22 7.34 55.22 0.00 8.55 6,695(%) 2004 9.14 7.40 49.64 0.00 8.32 1,657
2005 9.12 7.35 53.02 0.00 8.42 1,6722006 9.25 7.32 55.22 0.00 8.63 1,6802007 9.36 7.30 53.11 0.00 8.83 1,686
Foreign ALL 12.80 9.48 78.27 0.00 11.91 6,695(%) 2004 12.63 9.17 73.65 0.05 11.81 1,657
2005 12.55 9.22 77.63 0.05 11.68 1,6722006 13.18 9.97 77.45 0.00 12.07 1,6802007 12.84 9.42 78.27 0.00 12.07 1,686
Indiv ALL 32.10 29.46 95.98 1.81 17.06 6,695(%) 2004 32.90 30.57 93.51 2.90 16.76 1,657
2005 31.48 28.58 95.40 2.02 16.74 1,6722006 31.71 29.17 95.98 1.81 17.08 1,6802007 32.32 29.57 95.37 2.85 17.60 1,686
Note: Some of the minimum values of Tobin’s q (QRatio) are negative due to the negative equity capital of the
distressed firm.
30
Table 3 Correlations.
Panel A. Correlation between horizon, size, and turnover.
This table shows the correlation between HORIZON, TOP30, Log market value, and Turnover.
Here, HORIZON is a weighted average of the investment horizons of each firm (equation(2)),
and TOP30 is the sum of the top 30 shareholders’ holdings divided by the total number of
shares outstanding and Market value is the number of shares outstanding multiplied by the
stock price at the end of the fiscal year. Turnover is calculated as the daily trading volume
divided by the number of shares outstanding. The correlation coefficients are estimated
annually and averaged over the years 2004 to 2007.
HORIZON TOP30 Log Size TurnoverHORIZON 1 0.4631 0.2917 -0.1108
TOP30 0.4631 1 0.0130 -0.1347Log Size 0.2917 0.0130 1 0.0350Turnover -0.1108 -0.1347 0.0350 1.0000
Panel B. Correlation between HORIZON and investor category’s holding ratios.
This table shows the correlation between the HORIZON and the holding ratios of investor
categories such as foreign, cross-holding, and individual investors. The correlation coefficients
are estimated annually and averaged over the years 2004 to 2007.
HORIZON Foreign Crosshld IndivHORIZON 1 -0.0298 0.3493 -0.7557
Foreign -0.0298 1 -0.1221 -0.4885Crosshld 0.3493 -0.1221 1 -0.1224
Indiv -0.7557 -0.4885 -0.1224 1
31
Table 4 Panel least squares analysis of illiquidity.
This table shows the relation of illiquidity with investment horizon and the ownership
concentration ratio for stocks in the First and Second Sections of the TSE over the period 2004
to 2007. The results are from the panel least square regressions. The fixed period effect model
is selected as a result of the Hausman test. In order to avoid endogeneity problem, we use the
two-stage estimation with instrumental variables to obtain a consistent estimate. Our
instrumental variables are the lagged explanatory variables. RILLIQj,t is the relative ILLIQ
for firm j defined by equation (3), log_Sizej,t, is the natural logarithm of firm j’s market value
at the end of March, HORIZONj,t is the weighted average of the holding period for firm j’s
stockholders in year t, TOP30 is the sum of the top 30 shareholders’ holdings divided by the
total number of shares outstanding and Crosshldj,t and Foreignj,t are the percentages of
ownership of firm j’s investor categories at the end of fiscal year t. are
parameters to be estimated and
eanddcba ,,,,
is an error term. Heteroskedasticity is corrected by White
diagonal standard errors and covariance corrections.
Dependent Variable: RILLIQ
Model1 Model2 Model3
Coefficient t-value Coefficient t-value Coefficient t-value
HORIZON 0.1724 15.94 0.2751 17.60 0.2656 13.60
TOP30 0.7418 8.22
INSIDER 2.2483 12.97 0.0212 11.92
NON-INSIDER 0.3785 4.03
NON=INSIDER*Foreign -0.0046 -1.52
NON-INSIDER*Indiv 0.0051 1.42
NON=INSIDER*CrossHld 0.0159 5.38
Log Size -1.0736 -146.30 -1.0697 -143.50 -1.0416 -87.60
Intercept 2.0914 24.21 1.6658 16.87 1.6756 10.40
Adjusted R-squared 0.817 0.823 0.823
Observations 4,899 4,899 4,899
32
Table 5 Panel least squares analysis of Bid-Ask Spread
This table shows the relation of bid-ask spread with investment horizon and the ownership
concentration ratio for stocks in the First and Second Sections of the TSE over the period 2004
to 2007. The results are from the panel least square regressions. The fixed period effect model
is selected as a result of the Hausman test. In order to avoid endogeneity problem, we use the
two-stage estimation with instrumental variables to obtain a consistent estimate. Our
instrumental variables are the lagged explanatory variables. SPRD%j,t is time-weighted
quoted bid-ask spread divided by midprice, log_Sizej,t, is the natural logarithm of firm j’s
market value at the end of March, HORIZONj,t is the weighted average of the holding period
for firm j’s stockholders in year t, TOP30 is the sum of the top 30 shareholders’ holdings
divided by the total number of shares outstanding, and Crosshldj,t, Foreignj,t and Indiv j,t are
the percentages of ownership of firm j’s investor categories at the end of fiscal year t.
are parameters to be estimated and eanddcba ,,,, is an error term. Heteroskedasticity is
corrected by White diagonal standard errors and covariance corrections.
Dependent Variable: SPRD%
Model1 Model2 Model3
Coefficient t-value Coefficient t-value Coefficient t-value
HORIZON -0.0073 -1.17 0.0058 0.58 0.0045 0.39
TOP30 0.2463 6.00
INSIDER 0.4365 3.99 0.5741 4.65
NON-INSIDER 0.2079 4.68
NON=INSIDER*Foreign 0.0129 8.44
NON-INSIDER*Indiv -0.0087 -4.99
NON=INSIDER*CrossHld 0.0048 2.59
Log Trade -0.3588 -38.25 -0.3573 -39.79 -0.4031 -33.96
TIC/VWAP 1.0237 27.99 1.0376 26.02 1.0870 23.52
Intercept 4.0277 42.77 3.9523 45.89 4.5395 35.86
Adjusted R-squared 0.656 0.657 0.666
Observations 4,899 4,899 4,899
33
Table 6 Panel least squares analyses of Tobin’s q.
This table shows the relation between firm value and investment horizon, ownership
concentration, and ownership ratios for foreigners, individuals, and cross-holding for First-
and Second-Section stocks of the TSE over the period 2004 to 2007. To avoid any endogeneity
problems, we employ lagged variables as instrumental variables. The fixed period effect
model is selected as a result of the Hausman test. In order to avoid endogeneity problem, we
use the two-stage estimation with instrumental variables to obtain a consistent estimate. Our
instrumental variables are the lagged explanatory variables. Here, QRATIO is Tobin’s q =
(aggregate market value + interest-bearing debt)/(total capital + interest-bearing debt), where
the aggregate market value is adjusted for the cross-holding proportion of ownership. Foreign,
Indiv and Crosshld are the same as in equation (7). Debtasset is calculated as total debt
divided by total assets, and Growth is calculated as the one-year growth rate of the ordinary
profit which is equivalent to income before extraordinary items and taxes. Heteroskedasticity
is corrected by White diagonal standard errors and covariance corrections.
Dependent Variable: QRATIO
Model1 Model2 Model3 Model4
Coefficient t-value Coefficient t-value Coefficient t-value Coefficient t-value
HORIZON -0.2004 -7.09 -0.2130 -5.09 -0.0544 -1.53 -0.0839 -2.20
TOP30 0.0118 6.51
INSIDER 0.0100 2.69 0.0080 2.20 0.0068 1.73
NON-INSIDER 0.0122 5.73
NON=INSIDER*Foreign 0.0002 3.03 0.0001 2.56
NON-INSIDER*Indiv 0.0001 1.43
NON=INSIDER*CrossHld -0.0003 -6.39 -0.0003 -6.27
Debt/Asset -0.0044 -2.74 -0.0044 -2.73 -0.0027 -1.81 -0.0024 -1.72
Growth -1.2369 -2.28 -1.2390 -2.28 -1.1628 -2.26 -1.1564 -2.26
Log_Size 0.1990 12.83 0.1986 12.96 0.1397 5.19 0.1186 5.97
Intercept 1.0916 5.56 1.1441 4.58 0.9661 3.17 1.4066 5.32
F stats 134.6 118.2 100.5 111.5
Prob(F-statistic) 0.000 0.000 0.000 0.000
Observations 4,892 4,892 4,892 4,892
34
Table 7 The effects of changes in liquidity measures
This table shows the effects of changes in HORIZON and Top30 at a prior year to liquidity as
well as firm value. The results are from the panel least squares. In model1 where ⊿ILLIQ
is the change in ILLIQ for stock j from year t-1 to year t. In model2 ⊿SPRD% is the change
in percent bid-ask spread of stock j from year t-1 to year t. Changes in HORIZON, INSIDER
and NON-INSIDER at the previous year are included as explanatory variables. In model1, we
include the change in the market-average ILLIQ (Market_ILLIQ). In model2, we have ⊿Tic
and ⊿Price to adjust the effects from the TSE’s step-wise tick schedule. Definitions of
Debt/Asset and Growth are same as table 6. Heteroskedasticity is corrected by White diagonal
standard errors and covariance corrections.
Model1 Model2 Model3
Dependent Variable: ⊿ILLIQ ⊿SPRD% ⊿QRATIO
Coefficient t-value Coefficient t-value Coefficient t-value
⊿HORIZON(-1) -0.0570 -4.41 -0.0371 -3.59 -0.0038 -1.70
⊿INSIDER(-1) 0.0099 2.88 0.0067 3.13 0.0409 3.99
⊿NON-=INSIDER(-1) -0.0083 -2.32 0.0038 2.70 0.0017 4.97
⊿(Tic/Price) 0.9944 111.35 0.0061 0.93
⊿Trade -0.1251 -5.09
⊿Debt/Asset -0.0001 -6.58
⊿Growth -0.2346 -21.59
⊿Market_ILLIQ -32.6696 -8.18
Intercept 30.7375 8.22 0.2365 9.96 -0.2346 -21.59
Adjusted R-squared 0.046 0.329 0.126
Observations 3,181 3,181 3,181
35